This submission is designed in response to NIDA PAR-10-021 AIDS-Science Track Award for Research Transition (R03), titled The effects of client change language on adolescent risky sexual behavior. People under age 25, and those involved with the justice system in particular, are at heightened risk for sexually transmitted infections including HIV. Approaches to reduce sexual risk this population have had limited success, and it remains unclear whether adolescents can benefit from interventions that have been proven effective in adults. One approach that has shown promise is motivational interviewing (MI), a directive, client- centered method employed in the treatment of problematic health behaviors. The mechanism of action with the most empirical support in MI is change talk, spontaneous within-session commitments by clients to change their problem behavior. Clinicians trained in MI employ specific tactics to acknowledge, elicit, and strengthen this change talk, with the goal of increasing its frequency and strength. A causal chain has been established linking therapist speech, client change talk, and subsequent health outcomes in adults. This study will use the Motivational Interviewing Skill Code to quantify client change language from an ongoing study of criminally- involved adolescents. Participants in the parent study are randomized to either MI or a behavioral skills training (BST) condition, both of which have two intervention sessions one week apart, with follow-up assessments at three and six months. Coding client language in this study presents the opportunity to measure change language from a randomized controlled trial of two treatments: MI and BST. We will quantify client language in a sample of 196 treatment sessions, 98 from each condition. Outcome measures in the parent study include a time-line follow-back querying sexual behavior and substance use over the previous 30 days. Effects will be evaluated from both the 3- and 6-month follow-ups, with the maximal effect of treatment expected at 3 months. No previous study has successfully applied client language coding to the study of adolescent sexual risk. This secondary analysis will test a causal chain developed in adult populations by using an adolescent sample and comparing those effects in two treatment conditions, an approach that will guide clinical practice by either corroborating or challenging the effects of client change language in adolescents.